Researchers from Singapore recently evaluated the effectiveness of the app-based multidomain intervention (ADL+) for the prevention of cognitive decline in healthy community-dwelling older adults.
Participants received either a six-month intervention that included app-based cognitive training and personalized nutritional, physical and social activities recommendations or were assigned to a control group.
Results supported “the potential of digital multidomain lifestyle interventions for the prevention of cognitive decline,” the researchers wrote in the International Journal of Environmental Research and Public Health.
Technology-driven interventions for cognitive health
Nutrition plays a crucial role in cognitive health among older adults as deficiencies in essential nutrients have been linked to cognitive decline and dementia.
Poor nutrition can exacerbate memory loss, slow processing speed and reduce mental clarity. To address this, mobile apps are increasingly being used to monitor dietary intake, provide personalized nutrition recommendations and encourage healthier eating habits.
These apps provide personalized recommendations based on individual health goals, preferences and dietary restrictions. AI and gamification techniques can further personalize lifestyle interventions and encourage behavioral change, such as providing tailored nutrition and exercise recommendations.
However, evidence supporting technology-driven interventions remains limited and inconsistent, requiring further research to assess their clinical impact, the researchers noted.
AI toolkit helps older adults maintain memory and attention
The researchers recruited 142 community-dwelling older adults age 60 years and older without dementia but with subjective memory complaints.
Participants received either the intervention (ADL+ toolkit) or a control group treatment in the form of a cognitive health educational package.
The ADL+ toolkit was designed to actively monitor cognitive function and provide personalized interventions through seven interactive modules, integrating gamified cognitive training, physical exercise and diet tracking while analyzing user data to tailor interventions based on individual cognitive status.
The Diet Analysis and Recommendation module used AI and image-based technology to track food intake, analyze nutrition and encourage healthier eating. It employed advanced computer vision to identify food items from meal photos and estimate their nutritional values.
The developers created a data set of over 500 dishes and 140,000 images, and an artificial neural network analyzed images using a specialized nutrition database from the local Health Promotion Board, automatically retrieving and updating nutritional values.
The researchers evaluated cognitive outcomes using a standardized neuropsychological test battery (NTB) at baseline, six months and nine months.
They measured changes in NTB composite scores across attention, processing speed, memory and executive function domains, as well as activity levels and health-related quality of life.
At six months, the intervention group’s cognitive scores improved, while the control group’s scores declined. The intervention group performed better in attention, processing speed and memory but not in executive function.
By nine months, only the processing speed advantage remained, with no lasting cognitive improvements overall. The intervention did not improve cognitive function scores, but it may have prevented decline as seen in the control group.
The intervention group’s activity levels slightly increased at six months, while the control group’s levels dropped significantly. However, by nine months, both groups declined with no major differences.
Quality of life scores showed no significant differences between groups, but intervention participants reported feeling better overall.
“The beneficial effects observed on attention, memory, processing speed and overall neuropsychological test performance suggest that a personalized approach, facilitated by technology, may effectively deliver multidomain interventions,” the researchers wrote.
They noted that future research should explore strategies like continued platform engagement to maintain consistent levels of cognitive, physical and social activity for optimal outcomes.
“AI methods show potential for enhancing digital health interventions, but challenges such as computational demands, data annotation and explainability must be addressed for real-world implementation,” they concluded.
Source: Int. J. Environ. Res. Public Health 2025, 22(1), 42. doi: 10.3390/ijerph22010042. “ADL+: A Digital Toolkit for Multidomain Cognitive, Physical, and Nutritional Interventions to Prevent Cognitive Decline in Community-Dwelling Older Adults”. Authors: Chew, J. Et al.