AgeTech-Service Convergence Major, Department of Medical Nutrition, Graduate School of East-West Medical Science, Kyung Hee University, Yongin, Korea JFIF d d Ducky d qhttp://ns.adobe.com/xap/1.0/ |er˭|I|-fGK.rΛp8XV1%K6mvG+tc+qE &ǸC_Nm:l=_/m5^[dߌڇ.c<%:)tQ$Ow~-aY;UJ >=F)2[nk؆?훐 M=l6[4(O.]2#-H^n#->&mp5~Fӛ+|| S,xag% qkEUzUgæBhߕP(7]kFnq?֖Cpru Z6*rEڊtS|*tI*E}7R<,nUU֫^I7Q*mSly%rdȓd8hE<9oHhMfNSRj[i7D[Rj݊+kდq{"$$H?p \̅S?㭻;t~ R߁ )^/>Qj`yt[w ԛ;² ~+ ߔ_ YW~|o] ?x^ᯛ `ʼn;g)T@vWn]>&4l p+$D̢1l|ȨF%-} .9[}w~
ԠLM9hСablfe&QoW !s ?wjLK?s7yO>(=C~_nyǜu? v 3vyo oI@qV-jeES^[9WoSܝh"l2C1a͔CiJ@3:Pճw= /7ovuk+\V;lDgն <[A+r X~d;m!_s8ݖ;;.0llUC+?i#_crʙ1~C.\q u l8Hܶ2m`ܻM3Tov|Bs rɵ"oLS-
DКw=Tv@f '6|YlD͓Y%-#Ѯo%:&!3o%\J<02;K87>^vgƓ# ;ݝmz^Y6=PS39U%~
&f#
}o!muH; ʲŇ˷yvP+&.7e[3'vR4Yj̗IZ`e˽3o[WU{
m[sUbۋZǾۆl6~9'V*.\S2<Sd*zY[aŶ`]C$n.v^Ʌ
dng>ەZ,Mmϑ :n6nϦezW qUJ4! ۇ4R! =>>F n|Q[{pRO 17ƕ~._I ''00k=b՛o}Osðc2'o\3}ݭQ^ 2 . R1yKȣtAݿ-u ܾw!`?1Whn|gzU o[ECWwjUIן)^h#1ɭ!/Z
np;o;ΗŻkXs."6E`Z1
עӐ9Kl8qd q} 2Stt;#j>;խabONŗ=fwP1j)l6J̶|gV 2`y/0E˛6+ԫ1?
6}KW c\KoKͨ2ۅFws*TԞLיuDx .kCzWXhy۶gLu|%TnupǺl-S* P RaLnT+c+*xl.v!.U=|; !_L̎뱚U=4hm:ٯ"y)$:>%(n}X'p[ȴ
^˒4kƓmzDx
\
'Nq amP7nyN݅=
© 2026 The Korean Society of Clinical Nutrition
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 unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Author Contributions
Conceptualization: all authors. Methodology: all authors. Data curation: all authors. Investigation: all authors. Data extraction: SJ. Data synthesis and interpretation: all authors. Writing - original draft: SJ. Writing - review & editing: all authors. All authors read and approved the final manuscript.
Conflict of Interest
None.
Funding
None.
Data availability
Datasets are not publicly available but are available from the corresponding author upon reasonable request.
| Study | Country | Design | Population | Age (yr) | No. of participants | Duration | Setting | Digital technology | Delivery mode |
|---|---|---|---|---|---|---|---|---|---|
| Connelly et al. [23] 2025 | United Kingdom | Co-design/facilitation test | Care home older adults and staff | Older adults; not specified | 154 participants | 2-wk test | Care homes | Mobile app; AI system | AI-assisted assessment |
| Gaewkhiew et al. [24] 2025 | Thailand | Pilot RCT | Older dental patients | ≥60 y | 30 | 3 mon | Community | LINE app | Blended, dietitian-led |
| Su et al. [25 2023 | United States | Pilot RCT | Frail older adults | 66–77 y | 15 | 3 mon | Home-based | Mobile app (Olitor) | App-based feedback |
| Gutierrez et al.[26] 2025 | Chile | Iterative design + pilot | Community-dwelling older adults | Mean 71 y | 60 | 3 mon | Home | Mobile app (NeoMayor) | App-based self-monitoring |
| de Souto Barreto et al. [27] 2021 | France | Pilot RCT | Older adults with subjective memory complaints | ≥65 y | 120 | 6 mon | Community | Web platform/tablet | Web-based multidomain |
| Kurotani et al. [28] 2025 | Japan | Quasi-experimental | Community-dwelling frail older adults | ≥65 y | 29 | 3 mon | Community | AI-assisted smartphone app | AI feedback |
| Li et al. [29] 2025 | Taiwan | Quasi-experimental | Community-dwelling frail older adults | Mean 74 y | 81 | 6 mon | Community center/home | LINE app | Hybrid coaching |
| Gerber et al. [30] 2024 | United States | Pilot RCT | Overweight older adults at risk of diabetes | ≥60 y | 20 | Approximately ~6 months, 22 sessions | Online | Web/video system | Remote coach-led |
| Happe et al. [31] 2023 | Germany | Pilot study | Older rehabilitation inpatients | ≥70 y | 21 | 9 wk | Rehabilitation/home | Tablet app; ECA | E-coach |
| Rainero et al. [32] 2021 | Europe/Japan | RCT | Community-dwelling older adults at risk of frailty | Mean 70 y | 201 | 12 mon | Home | ICT platform, sensors, and app | Algorithm-based package |
| van den Helder et al. [33] 2020 | Netherlands | Cluster RCT | Community-dwelling older adults | Mean 72 y | 224 | 6 mon | Community/home | Tablet app (VITAMIN) | Blended app + expert coaching |
| Pei-Shan Li et al. [34] 2025 | Taiwan | Quasi-experimental | Community-dwelling frail/pre-frail older adults | ≥65 y | 81 | 6 mon | Community care center/home | LINE app | Hybrid coaching |
| Chew et al. [35] 2025 | Singapore | Quasi-experimental | Community-dwelling older adults at risk of cognitive decline | ≥60 y | 150 | 6 mon | Senior center/app | AI toolkit, mobile app, Kinect | AI-supported multidomain toolkit |
| Study | Objectives | Intervention description | Outcome: efficacy | Outcome: users’ perceptions |
|---|---|---|---|---|
| Connelly et al. [23] (2025) | To design and evaluate a digital nutrition assessment tool for identifying food preferences and malnutrition risk among older adults in care homes. | A mobile app and AI-supported system were used for meal photo recording, preference documentation, symptom monitoring, and malnutrition risk alerts using care-home menu data. | The system demonstrated potential to alleviate documentation burden and support malnutrition risk identification through AI-assisted food image recognition. | Older adults and care staff perceived the tool as useful, particularly because image-based input reduced manual recording burden. Limited food database coverage was noted. |
| Gaewkhiew et al. [24] (2025) | To examine the impact of nutrition counseling delivered via an online application on dietary intake among older adults. | The participants received individualized nutrition counseling through LINE based on 3-day dietary records, health conditions, and eating habits. | Retention was 100%. Energy and protein intake exhibited improvement trends, and nutritional status assessed by MNA-SF was monitored. | LINE was considered familiar and accessible for older adults, supporting the feasibility of remote expert-led nutrition counseling. |
| Su et al. [25] (2023) | To evaluate a mobile intervention designed to support healthy eating among frail older adults. | The Olitor mobile app supported Mediterranean diet recording and provided visual feedback by comparing individual MedD scores with recommended levels. | Adherence to the Mediterranean diet significantly improved in the intervention group. Physical function indicators, such as SPPB, were also assessed. | The small pilot study suggested that visual feedback was acceptable and could motivate dietary behavior change among frail older adults. |
| Gutierrez et al. [26] (2025) | To develop and pilot-test the NeoMayor mHealth app for promoting healthy lifestyles among older adults in Chile. | The app provided lifestyle recommendations based on Life’s Essential 8, including nutrition-related feedback, self-monitoring, and personalized images/messages based on baseline health scores. | Cardiovascular health scores significantly improved, as well as diet quality indicators, including the MIND score, following the intervention. | The participants generally accepted the app-based self-monitoring approach, although the study sample was predominantly female and relied on self-reported data. |
| de Souto Barreto et al. [27] (2021) | To evaluate the feasibility of a web-based multidomain lifestyle intervention for older adults with subjective memory complaints. | A web platform or tablet delivered multidomain lifestyle content, including cognitive training, physical activity, and nutrition education, with regular online contact. | The intervention demonstrated good feasibility and acceptability. Cognitive function and nutritional status indicators, including MNA, were evaluated. | Regular online contact and structured web content were perceived as feasible, although imbalance in participant characteristics limited the interpretation. |
| Kurotani et al. [28] (2025) | To determine whether an AI-assisted smartphone application could improve diet quality among community-dwelling frail older adults. | The app enabled recording of 10 food groups and provided AI-assisted personalized feedback using image analysis and text-based recommendations. | The intervention group demonstrated a significant improvement in the dietary balance score compared with the control group. | The AI-assisted feedback was considered useful for real-time dietary self-management, although the study was small and nonrandomized. |
| Li et al. [29] (2025) | To evaluate the effects of the Fitness and Nutrition Program for Seniors on frailty prevention and physical function among older adults. | A hybrid intervention combined face-to-face sessions with LINE-based remote coaching, nutrition education, and monitoring of dietary behavior. | The SPPB score and 4-m gait speed significantly improved. BMI and calf circumference were maintained during the intervention. | The hybrid model was feasible during the COVID-19 pandemic and supported continued engagement through familiar mobile communication. |
| Gerber et al. [30] (2024) | To evaluate the feasibility of an eHealth diabetes prevention program adapted for older adults at risk of diabetes. | A web-based/video-delivered lifestyle intervention based on the Diabetes Prevention Program employed remote coaching, behavioral strategies, and individualized progress reports. | Participation was high, and body weight exhibited a favorable decreasing trend. Adherence and session completion were also evaluated. | The participants demonstrated high engagement with the remote program, supporting the feasibility of counselor-supported eHealth coaching for older adults. |
| Happe et al. [31] (2023) | To evaluate the usability and feasibility of a tablet-based e-coach for older adults in rehabilitation. | A tablet app with an embodied conversational agent supported nutrition and exercise diary, stage-based goal setting, and tailored guidance based on the transtheoretical model. | The average SUS score was favorable, and adherence to nutrition-related recommendations improved during the intervention. | The e-coach format was perceived as usable and supportive. However, the study had no control group and experienced attrition. |
| Rainero et al. [32] (2021) | To determine whether the My-AHA ICT platform could prevent declines in quality of life among older adults at risk of frailty. | An ICT platform integrating apps and sensors provided multidomain interventions, including nutrition, cognition, physical activity, and sleep management, matched to individual needs. | The intervention helped prevent declines in quality of life compared with the control group. Nutritional status indicators, such as MNA, were also monitored. | The platform demonstrated feasibility for integrating multidomain risk assessment and individualized intervention delivery in older adults. |
| van den Helder et al. [33] (2020) | To evaluate the effects of blended home-based exercise and dietary protein intervention on physical function in community-dwelling older adults. | A tablet-supported program combined resistance exercise with dietary protein counseling. App-recorded dietary data were utilized to facilitate expert nutrition coaching. | Protein intake significantly increased in the exercise-plus-protein group. Physical performance outcomes, including m-PPT, were evaluated. | The blended approach supported home-based coaching, although the high dropout rate and nonblinded design were noted as limitations. |
| Li et al. [34] (2025) | To determine whether the FANS program improved frailty status and health-related quality of life among frail or pre-frail older adults. | The program used community center sessions and LINE-based home care. Monthly individualized home and group tasks were adjusted according to behavioral stage and cognitive-behavioral strategies. | The frailty scores and instrumental activities of daily living significantly improved compared with the control group. More than 80% of the participants reached the maintenance stage for dietary behavioral change. | The hybrid intervention was considered practical for frail older adults and supported sustained behavioral change through stage-based personalization. |
| Chew et al. [35] (2025) | To assess the ADL+ digital toolkit for multidomain cognitive, physical, and nutritional interventions among older adults at risk of cognitive decline. | The AI-based mobile toolkit utilized image-based dietary records, AI nutrition analysis, Kinect-based activities, and cognitive prediction models to adjust intervention intensity. | Global cognitive scores improved in the intervention group, particularly in processing speed. Activity levels were maintained in the intervention group compared with the control group. | Gamified and AI-supported features promoted engagement and social interaction; however, differences in digital literacy may have influenced cognitive burden. |
References 29 and 34 are companion reports derived from the same FANS study cohort and were counted as one unique study.
AI, artificial intelligence; MNA-SF, Mini Nutritional Assessment–Short Form; MedD, Mediterranean diet; SPPB, short physical performance battery; COVID-19, coronavirus disease 2019; SUS, System Usability Scale; ICT, information and communication technology; m-PPT, modified Physical Performance Test; FANS, fitness and nutrition program for seniors; ADL, activity of daily living.
| Outcome variable | Study | Results |
|---|---|---|
| Dietary intake, diet quality, and dietary adherence | Gaewkhiew et al. [24] (2025) | No between-group differences in nutrient intake; both groups showed decreased energy/nutrient intake over time (P<0.05) |
| Su et al. [25] (2023) | Mediterranean diet adherence improved (adjusted P=0.04); legume intake increased (adjusted P<0.01) | |
| Kurotani et al. [28] (2025) | Dietary quality improved (+9.5 points; 95% CI, 2.3–16.7; P=0.01) | |
| van den Helder et al. [33] (2020) | Protein intake increased (+0.32 g/kg/day, P<0.001) | |
| Nutritional status, malnutrition risk, and anthropometric outcomes | Connelly et al. [23] (2025) | Malnutrition risk and food preference identification supported. P-value not reported |
| Gaewkhiew et al. [24] (2025) | No between-group difference in body measurements | |
| Gutierrez et al. [26] (2025) | Waist circumference decreased, and CVH index improved (P<0.001) | |
| Li et al. [29] (2025) | Mid-upper arm circumference increased (P<0.001); calf circumference improved (P<0.01) | |
| Physical function, frailty, and daily functioning | Li et al. [29] (2025) | SPPB improved (P<0.05); Five Times Sit-to-Stand test performance improved (P<0.05); TUG test performance improved (P<0.001); physical activity level increased (P<0.001) |
| van den Helder et al. [33] (2020) | m-PPT, no difference (HBex, P=0.933; HBex-Pro, P=0.730); gait speed, improved (P=0.001); muscle strength, improved (P=0.001); muscle mass, improved (P=0.017) | |
| Li et al. [34] (2025) | Frailty and daily activities improved; exact P-value not reported in the abstract | |
| Cognitive and brain-related outcomes | Chew et al. [35] (2024) | NTB composite score improved (between-group difference=0.17; 95% CI, 0.071–0.27; P=0.001) |
| Rainero et al. [32] (2021) | Stroop C incongruent time improved (P=0.012); stroop C incongruent errors improved (P=0.015) | |
| de Souto Barreto et al. [27] (2021) | No significant effect for most cognitive outcomes | |
| Cardiometabolic and health-related indicators | Gutierrez et al. [26] (2025) | CVH index improved from 64 to 68 (P<0.001) |
| Gerber et al. [30] (2024) | Intervention group lost 9.5% body weight; control group gained 2.4%; 90% lost ≥5% body weight | |
| Su et al. [25] (2023) | HOMA-IR, no difference (adjusted P=0.85) | |
| Behavioral change, engagement, and maintenance | Gaewkhiew et al. [24] (2025) | Retention, 100% |
| Su et al. [25] (2023) | Retention, 100%; average app use, about 12 min/wk | |
| Gerber et al. [30] (2024) | Attendance and self-monitoring adherence, 100%; greater weight loss among eating behavioral change participants (−12.0% vs. −5.0%, P<0.005) | |
| Happe et al. [31] (2023) | Nutrition recommendation achievement increased from 24% to 66%; physical activity recommendation achievement increased from 33% to 71%. P-value not reported | |
| Li et al. [29] (2025) | Dietary behavioral action/maintenance stage reached by 81.4% | |
| Usability, acceptability, and users’ perceptions | Connelly et al. [23] (2025) | AI-assisted food image recording reduced documentation burden. P-value not reported |
| Happe et al. [31] (2023) | SUS, 78.6 | |
| Li et al. [29] (2025) | Satisfaction, 9.25/10 | |
| Chew et al. [35] (2024) | High adherence; average usability |
CI, confidence interval; CVH, cardiovascular health; SPPB, Short Physical Performance Battery; TUG, Timed Up and Go; m-PPT, modified Physical Performance Test; HBex, home-based exercise; HBex-Pro, home-based exercise plus protein; HOMA-IR, homeostatic model assessment of insulin resistance; SUS, System Usability Scale.
| Category | Inclusion criteria | Exclusion criteria |
|---|---|---|
| Population | Studies targeting older adults aged ≥60 years and reviewing community-dwelling, institutionalized, frail, prefrail, nutritionally at risk, or sarcopenia-risk older adults, as well as older adults with chronic conditions | Studies targeting children, adolescents, younger adults, or mixed-age populations that did not separately identify data for older adults |
| Concept: nutrition intervention | Studies focusing on nutrition education, counseling, dietary feedback, meal planning, dietary behavior change, nutrition monitoring, or malnutrition prevention/management | Studies without nutrition or dietary intervention components |
| Concept: digital or AI-enabled component | Studies using apps, web platforms, telehealth, messaging tools, ICT platforms, wearable devices, algorithms, AI, or image-based dietary assessment to implement or support the intervention | Face-to-face-only interventions without digital, remote, ICT, or AI-enabled components |
| Concept: personalization strategy | Studies designing the intervention based on dietary intake, nutritional status, health status, preferences, physical function, frailty, behavior, user profile, app data, wearable data, or AI-generated recommendations | Studies providing only standardized general information without individualized feedback, tailoring, goal setting, or recommendations |
| Study type | RCTs, nonrandomized trials, quasi-experimental studies, pilot/feasibility studies, mixed-methods studies, and observational intervention studies | Reviews, editorials, commentaries, letters, conference abstracts, theses, animal studies, cell studies, and non-intervention studies |
| Publication type | Peer-reviewed journal articles | Gray literature and non-peer-reviewed documents |
| Language | English | Non-English articles |
| Publication year | Articles published from January 1, 2010, through June 30, 2026 | Articles published before January 1, 2010, or after the final search date |
| PCC element | Definition in this review |
|---|---|
| Population (P) | Older adults aged ≥60 years, including community-dwelling, institutionalized, frail, prefrail, and nutritionally at risk older adults |
| Concept (C) | Personalization strategies in digital or AI-enabled nutrition interventions, such as individualized dietary feedback, tailored nutrition counseling, meal planning, self-monitoring, and algorithm- or AI-supported recommendations |
| Context (C) | Community, home-based, clinical, institutional, long-term care, and remote or telehealth-based settings |
| Study | Country | Design | Population | Age (yr) | No. of participants | Duration | Setting | Digital technology | Delivery mode |
|---|---|---|---|---|---|---|---|---|---|
| Connelly et al. [23] 2025 | United Kingdom | Co-design/facilitation test | Care home older adults and staff | Older adults; not specified | 154 participants | 2-wk test | Care homes | Mobile app; AI system | AI-assisted assessment |
| Gaewkhiew et al. [24] 2025 | Thailand | Pilot RCT | Older dental patients | ≥60 y | 30 | 3 mon | Community | LINE app | Blended, dietitian-led |
| Su et al. [25 2023 | United States | Pilot RCT | Frail older adults | 66–77 y | 15 | 3 mon | Home-based | Mobile app (Olitor) | App-based feedback |
| Gutierrez et al.[26] 2025 | Chile | Iterative design + pilot | Community-dwelling older adults | Mean 71 y | 60 | 3 mon | Home | Mobile app (NeoMayor) | App-based self-monitoring |
| de Souto Barreto et al. [27] 2021 | France | Pilot RCT | Older adults with subjective memory complaints | ≥65 y | 120 | 6 mon | Community | Web platform/tablet | Web-based multidomain |
| Kurotani et al. [28] 2025 | Japan | Quasi-experimental | Community-dwelling frail older adults | ≥65 y | 29 | 3 mon | Community | AI-assisted smartphone app | AI feedback |
| Li et al. [29] 2025 | Taiwan | Quasi-experimental | Community-dwelling frail older adults | Mean 74 y | 81 | 6 mon | Community center/home | LINE app | Hybrid coaching |
| Gerber et al. [30] 2024 | United States | Pilot RCT | Overweight older adults at risk of diabetes | ≥60 y | 20 | Approximately ~6 months, 22 sessions | Online | Web/video system | Remote coach-led |
| Happe et al. [31] 2023 | Germany | Pilot study | Older rehabilitation inpatients | ≥70 y | 21 | 9 wk | Rehabilitation/home | Tablet app; ECA | E-coach |
| Rainero et al. [32] 2021 | Europe/Japan | RCT | Community-dwelling older adults at risk of frailty | Mean 70 y | 201 | 12 mon | Home | ICT platform, sensors, and app | Algorithm-based package |
| van den Helder et al. [33] 2020 | Netherlands | Cluster RCT | Community-dwelling older adults | Mean 72 y | 224 | 6 mon | Community/home | Tablet app (VITAMIN) | Blended app + expert coaching |
| Pei-Shan Li et al. [34] 2025 | Taiwan | Quasi-experimental | Community-dwelling frail/pre-frail older adults | ≥65 y | 81 | 6 mon | Community care center/home | LINE app | Hybrid coaching |
| Chew et al. [35] 2025 | Singapore | Quasi-experimental | Community-dwelling older adults at risk of cognitive decline | ≥60 y | 150 | 6 mon | Senior center/app | AI toolkit, mobile app, Kinect | AI-supported multidomain toolkit |
| Study | Objectives | Intervention description | Outcome: efficacy | Outcome: users’ perceptions |
|---|---|---|---|---|
| Connelly et al. [23] (2025) | To design and evaluate a digital nutrition assessment tool for identifying food preferences and malnutrition risk among older adults in care homes. | A mobile app and AI-supported system were used for meal photo recording, preference documentation, symptom monitoring, and malnutrition risk alerts using care-home menu data. | The system demonstrated potential to alleviate documentation burden and support malnutrition risk identification through AI-assisted food image recognition. | Older adults and care staff perceived the tool as useful, particularly because image-based input reduced manual recording burden. Limited food database coverage was noted. |
| Gaewkhiew et al. [24] (2025) | To examine the impact of nutrition counseling delivered via an online application on dietary intake among older adults. | The participants received individualized nutrition counseling through LINE based on 3-day dietary records, health conditions, and eating habits. | Retention was 100%. Energy and protein intake exhibited improvement trends, and nutritional status assessed by MNA-SF was monitored. | LINE was considered familiar and accessible for older adults, supporting the feasibility of remote expert-led nutrition counseling. |
| Su et al. [25] (2023) | To evaluate a mobile intervention designed to support healthy eating among frail older adults. | The Olitor mobile app supported Mediterranean diet recording and provided visual feedback by comparing individual MedD scores with recommended levels. | Adherence to the Mediterranean diet significantly improved in the intervention group. Physical function indicators, such as SPPB, were also assessed. | The small pilot study suggested that visual feedback was acceptable and could motivate dietary behavior change among frail older adults. |
| Gutierrez et al. [26] (2025) | To develop and pilot-test the NeoMayor mHealth app for promoting healthy lifestyles among older adults in Chile. | The app provided lifestyle recommendations based on Life’s Essential 8, including nutrition-related feedback, self-monitoring, and personalized images/messages based on baseline health scores. | Cardiovascular health scores significantly improved, as well as diet quality indicators, including the MIND score, following the intervention. | The participants generally accepted the app-based self-monitoring approach, although the study sample was predominantly female and relied on self-reported data. |
| de Souto Barreto et al. [27] (2021) | To evaluate the feasibility of a web-based multidomain lifestyle intervention for older adults with subjective memory complaints. | A web platform or tablet delivered multidomain lifestyle content, including cognitive training, physical activity, and nutrition education, with regular online contact. | The intervention demonstrated good feasibility and acceptability. Cognitive function and nutritional status indicators, including MNA, were evaluated. | Regular online contact and structured web content were perceived as feasible, although imbalance in participant characteristics limited the interpretation. |
| Kurotani et al. [28] (2025) | To determine whether an AI-assisted smartphone application could improve diet quality among community-dwelling frail older adults. | The app enabled recording of 10 food groups and provided AI-assisted personalized feedback using image analysis and text-based recommendations. | The intervention group demonstrated a significant improvement in the dietary balance score compared with the control group. | The AI-assisted feedback was considered useful for real-time dietary self-management, although the study was small and nonrandomized. |
| Li et al. [29] (2025) | To evaluate the effects of the Fitness and Nutrition Program for Seniors on frailty prevention and physical function among older adults. | A hybrid intervention combined face-to-face sessions with LINE-based remote coaching, nutrition education, and monitoring of dietary behavior. | The SPPB score and 4-m gait speed significantly improved. BMI and calf circumference were maintained during the intervention. | The hybrid model was feasible during the COVID-19 pandemic and supported continued engagement through familiar mobile communication. |
| Gerber et al. [30] (2024) | To evaluate the feasibility of an eHealth diabetes prevention program adapted for older adults at risk of diabetes. | A web-based/video-delivered lifestyle intervention based on the Diabetes Prevention Program employed remote coaching, behavioral strategies, and individualized progress reports. | Participation was high, and body weight exhibited a favorable decreasing trend. Adherence and session completion were also evaluated. | The participants demonstrated high engagement with the remote program, supporting the feasibility of counselor-supported eHealth coaching for older adults. |
| Happe et al. [31] (2023) | To evaluate the usability and feasibility of a tablet-based e-coach for older adults in rehabilitation. | A tablet app with an embodied conversational agent supported nutrition and exercise diary, stage-based goal setting, and tailored guidance based on the transtheoretical model. | The average SUS score was favorable, and adherence to nutrition-related recommendations improved during the intervention. | The e-coach format was perceived as usable and supportive. However, the study had no control group and experienced attrition. |
| Rainero et al. [32] (2021) | To determine whether the My-AHA ICT platform could prevent declines in quality of life among older adults at risk of frailty. | An ICT platform integrating apps and sensors provided multidomain interventions, including nutrition, cognition, physical activity, and sleep management, matched to individual needs. | The intervention helped prevent declines in quality of life compared with the control group. Nutritional status indicators, such as MNA, were also monitored. | The platform demonstrated feasibility for integrating multidomain risk assessment and individualized intervention delivery in older adults. |
| van den Helder et al. [33] (2020) | To evaluate the effects of blended home-based exercise and dietary protein intervention on physical function in community-dwelling older adults. | A tablet-supported program combined resistance exercise with dietary protein counseling. App-recorded dietary data were utilized to facilitate expert nutrition coaching. | Protein intake significantly increased in the exercise-plus-protein group. Physical performance outcomes, including m-PPT, were evaluated. | The blended approach supported home-based coaching, although the high dropout rate and nonblinded design were noted as limitations. |
| Li et al. [34] (2025) | To determine whether the FANS program improved frailty status and health-related quality of life among frail or pre-frail older adults. | The program used community center sessions and LINE-based home care. Monthly individualized home and group tasks were adjusted according to behavioral stage and cognitive-behavioral strategies. | The frailty scores and instrumental activities of daily living significantly improved compared with the control group. More than 80% of the participants reached the maintenance stage for dietary behavioral change. | The hybrid intervention was considered practical for frail older adults and supported sustained behavioral change through stage-based personalization. |
| Chew et al. [35] (2025) | To assess the ADL+ digital toolkit for multidomain cognitive, physical, and nutritional interventions among older adults at risk of cognitive decline. | The AI-based mobile toolkit utilized image-based dietary records, AI nutrition analysis, Kinect-based activities, and cognitive prediction models to adjust intervention intensity. | Global cognitive scores improved in the intervention group, particularly in processing speed. Activity levels were maintained in the intervention group compared with the control group. | Gamified and AI-supported features promoted engagement and social interaction; however, differences in digital literacy may have influenced cognitive burden. |
| Outcome variable | Study | Results |
|---|---|---|
| Dietary intake, diet quality, and dietary adherence | Gaewkhiew et al. [24] (2025) | No between-group differences in nutrient intake; both groups showed decreased energy/nutrient intake over time (P<0.05) |
| Su et al. [25] (2023) | Mediterranean diet adherence improved (adjusted P=0.04); legume intake increased (adjusted P<0.01) | |
| Kurotani et al. [28] (2025) | Dietary quality improved (+9.5 points; 95% CI, 2.3–16.7; P=0.01) | |
| van den Helder et al. [33] (2020) | Protein intake increased (+0.32 g/kg/day, P<0.001) | |
| Nutritional status, malnutrition risk, and anthropometric outcomes | Connelly et al. [23] (2025) | Malnutrition risk and food preference identification supported. P-value not reported |
| Gaewkhiew et al. [24] (2025) | No between-group difference in body measurements | |
| Gutierrez et al. [26] (2025) | Waist circumference decreased, and CVH index improved (P<0.001) | |
| Li et al. [29] (2025) | Mid-upper arm circumference increased (P<0.001); calf circumference improved (P<0.01) | |
| Physical function, frailty, and daily functioning | Li et al. [29] (2025) | SPPB improved (P<0.05); Five Times Sit-to-Stand test performance improved (P<0.05); TUG test performance improved (P<0.001); physical activity level increased (P<0.001) |
| van den Helder et al. [33] (2020) | m-PPT, no difference (HBex, P=0.933; HBex-Pro, P=0.730); gait speed, improved (P=0.001); muscle strength, improved (P=0.001); muscle mass, improved (P=0.017) | |
| Li et al. [34] (2025) | Frailty and daily activities improved; exact P-value not reported in the abstract | |
| Cognitive and brain-related outcomes | Chew et al. [35] (2024) | NTB composite score improved (between-group difference=0.17; 95% CI, 0.071–0.27; P=0.001) |
| Rainero et al. [32] (2021) | Stroop C incongruent time improved (P=0.012); stroop C incongruent errors improved (P=0.015) | |
| de Souto Barreto et al. [27] (2021) | No significant effect for most cognitive outcomes | |
| Cardiometabolic and health-related indicators | Gutierrez et al. [26] (2025) | CVH index improved from 64 to 68 (P<0.001) |
| Gerber et al. [30] (2024) | Intervention group lost 9.5% body weight; control group gained 2.4%; 90% lost ≥5% body weight | |
| Su et al. [25] (2023) | HOMA-IR, no difference (adjusted P=0.85) | |
| Behavioral change, engagement, and maintenance | Gaewkhiew et al. [24] (2025) | Retention, 100% |
| Su et al. [25] (2023) | Retention, 100%; average app use, about 12 min/wk | |
| Gerber et al. [30] (2024) | Attendance and self-monitoring adherence, 100%; greater weight loss among eating behavioral change participants (−12.0% vs. −5.0%, P<0.005) | |
| Happe et al. [31] (2023) | Nutrition recommendation achievement increased from 24% to 66%; physical activity recommendation achievement increased from 33% to 71%. P-value not reported | |
| Li et al. [29] (2025) | Dietary behavioral action/maintenance stage reached by 81.4% | |
| Usability, acceptability, and users’ perceptions | Connelly et al. [23] (2025) | AI-assisted food image recording reduced documentation burden. P-value not reported |
| Happe et al. [31] (2023) | SUS, 78.6 | |
| Li et al. [29] (2025) | Satisfaction, 9.25/10 | |
| Chew et al. [35] (2024) | High adherence; average usability |
| Identified theme | Description |
|---|---|
| Acceptability of digital nutrition interventions | Several studies showed that digital nutrition interventions were generally acceptable to older adults when the tools were simple, familiar, and directly related to their daily dietary practices. Mobile apps, web-based platforms, and messaging applications, such as LINE, were considered feasible delivery methods, particularly when supporting dietary self-monitoring, nutrition education, or remote counseling. |
| Reduced burden through automated or simplified dietary recording | Image-based dietary recording, AI-assisted food recognition, and structured app-based food logs alleviated the burden of manual dietary recording. These approaches were specifically relevant for older adults and care staff as they reduced the effort required to document meals and supported more efficient nutrition monitoring. |
| Importance of personalization and feedback | Personalized feedback based on dietary intake, nutritional status, health risk, behavioral stage, food preference, or activity data was considered useful for supporting behavioral changes. Studies using visual feedback, individualized goals, expert counseling, or algorithm-based recommendations reported that personalization may enhance engagement and adherence to nutrition-related behaviors. |
| Role of human support in digital interventions | Although digital tools were widely used, several interventions relied on nutritionists, counselors, coaches, or care staff to interpret data and offer individualized guidance. Blended models combining digital platforms with professional support appeared particularly useful for older adults, who may require additional encouragement, clarification, or motivational support. |
| Digital literacy and usability challenges | Some studies showed barriers associated with low digital literacy, unfamiliarity with digital platforms, complex navigation, or difficulties using digital devices. Older adults preferred clear instructions, simple interfaces, readable visual materials, and direct messages. These findings underscore the need for age-friendly design in digital nutrition interventions. |
| Engagement and behavior maintenance | Interventions including goal setting, self-monitoring, regular contact, tailored tasks, or stage-based behavioral change strategies appeared to support continued engagement. Structured digital follow-up and ongoing professional contact may help older adults sustain nutrition-related behaviors. |
| Feasibility of AI-enabled approaches | AI-enabled approaches, such as image recognition, algorithm-based recommendations, and generative AI feedback, potentially supported real-time and individualized dietary guidance. However, AI use remained limited across studies, and issues such as food database coverage, food recognition accuracy, small sample sizes, and validation of AI-generated recommendations should be considered. |
| Need for culturally and contextually appropriate design | Included studies revealed that digital nutrition interventions should consider care settings, user preferences, digital literacy, and available resources. Familiar communication tools and practical support from healthcare professionals or care staff may improve the relevance and usability of interventions for older adults. |
AI, artificial intelligence; ICT, information and communications technology; RCT, randomized controlled trial.
AI, artificial intelligence.
References 29 and 34 are companion reports from the same FANS study cohort and were counted as one unique study.
AI, artificial intelligence; ECA, embodied conversational agent; ICT, information and communication technology; RCT, randomized controlled trial.
References 29 and 34 are companion reports derived from the same FANS study cohort and were counted as one unique study.
AI, artificial intelligence; MNA-SF, Mini Nutritional Assessment–Short Form; MedD, Mediterranean diet; SPPB, short physical performance battery; COVID-19, coronavirus disease 2019; SUS, System Usability Scale; ICT, information and communication technology; m-PPT, modified Physical Performance Test; FANS, fitness and nutrition program for seniors; ADL, activity of daily living.
CI, confidence interval; CVH, cardiovascular health; SPPB, Short Physical Performance Battery; TUG, Timed Up and Go; m-PPT, modified Physical Performance Test; HBex, home-based exercise; HBex-Pro, home-based exercise plus protein; HOMA-IR, homeostatic model assessment of insulin resistance; SUS, System Usability Scale.
AI, artificial intelligence.
