Dementia, MCI
Conditions
Keywords
COM-B, artificial intelligence, digital health intervention, mHealth, conversational agent, large language model, Whatsapp-based intervention, Just-in-time adaptive intervention (JITAI), brain health coaching, behavioural nudging
Brief summary
The goal of this clinical trial is to evaluate whether an AI-powered digital brain health coaching intervention can improve modifiable brain health risk factors in community-dwelling adults aged 40 and above in Singapore. The main question it aims to answer is: does the use of BLAZE, an AI-powered digital brain health coaching intervention, lead to a greater improvement in Brain Care Score (BCS) compared to usual care at 6 months? Researchers will compare participants receiving BLAZE to those receiving usual care (waitlist control) to see if BLAZE leads to greater improvements in brain health risk factors, behavioural outcomes, quality of life, and cost-effectiveness. Participants will: * Be randomly assigned to either receive BLAZE immediately for 12 months or be placed on a waitlist receiving usual care for the first 6 months before receiving BLAZE for a further 6 months * Complete a series of assessments at baseline, 6 months, and 12 months including the Brain Care Score, physical function tests, cognitive assessments, quality of life measures, and lifestyle behaviour questionnaires * If assigned to the intervention group, interact with BLAZE via a WhatsApp-based platform that delivers personalised brain health recommendations, behavioural nudges, educational content, and goal-setting support based on their individual Brain Care Score profile
Detailed description
Dementia represents a growing public health challenge in Singapore and across the Asia-Pacific region, with prevalence projected to rise substantially as the population ages. Up to 40% of dementia risk is attributable to modifiable lifestyle factors, highlighting significant potential for prevention through targeted behavioural interventions. Despite this, translating evidence-based lifestyle recommendations into sustained behavioural change in community settings remains a significant challenge, particularly given the limited time available for personalised guidance in primary care. The THINK-SHARP trial evaluates BLAZE (Brain health Lifestyle Action with personaliZed Engagement), a theory-driven AI-powered digital coaching intervention designed to support brain health behaviours in community-dwelling adults. BLAZE is grounded in the Capability-Opportunity-Motivation-Behaviour (COM-B) framework, a well-established behavioural science model, and is delivered via a WhatsApp-based platform integrated with a large language model (LLM). The intervention operates through four integrated modules - Assessment, Education, Nudge, and Coaching - to deliver personalised recommendations, behavioural prompts, and goal-setting support tailored to each participant's individual brain health profile. Personalisation is driven by the Brain Care Score (BCS), a validated composite measure of modifiable physical, lifestyle, and social-emotional risk factors for brain health. BLAZE uses each participant's BCS profile to identify priority areas for behavioural change and generates adaptive, data-driven recommendations accordingly. All content is informed by a curated knowledge base developed in collaboration with neurologists and geriatricians to ensure clinical relevance and safety. The trial employs a two-arm, waitlist-controlled randomised controlled trial (RCT) design, with 350 participants randomised 1:1 to either the immediate intervention arm or the waitlist control arm. The immediate intervention arm receives BLAZE for 12 months, while the waitlist control arm receives usual care for the first 6 months before crossing over to receive BLAZE for a further 6 months. The primary outcome is the between-group difference in BCS at 6 months. Secondary outcomes assess changes across six domains: physical function, lifestyle behaviours, self-management and health engagement, cognitive function, quality of life, and social health. Beyond clinical effectiveness, the trial incorporates a concurrent process evaluation to assess implementation outcomes including adherence, feasibility, usability, and perceived value of the intervention. A model-based health economic evaluation will also be conducted to estimate the long-term cost-effectiveness of BLAZE from the Singapore public healthcare payer perspective. Together, these components provide a comprehensive assessment of BLAZE's potential as a scalable, behaviourally grounded approach to community brain health promotion. This trial also contributes novel evidence on the applicability of the Brain Care Score within a Southeast Asian community-dwelling population, extending its validation beyond previously studied settings and supporting its relevance for regional brain health promotion efforts.
Interventions
BEHAVIORALBLAZE (Brain health Lifestyle Action with personaliZed Engagement)
BLAZE is an AI-powered digital brain health coaching intervention delivered via a WhatsApp-based platform integrated with a large language model (LLM). Grounded in the COM-B behavioural framework, BLAZE operates through four integrated modules - Assessment, Education, Nudge, and Coaching - to deliver personalised recommendations, behavioural nudges, educational content, and goal-setting support. Personalisation is driven by each participant's Brain Care Score (BCS) profile, with intervention intensity stratified according to domain-specific scores. All content is informed by a curated knowledge base developed in collaboration with neurologists and geriatricians
Sponsors
Alexandra Hospital
National University Hospital, Singapore
National University Polyclinics, Singapore
Study design
Allocation
RANDOMIZED
Intervention model
SEQUENTIAL
Primary purpose
HEALTH_SERVICES_RESEARCH
Masking
SINGLE (Outcomes Assessor)
Intervention model description
This study is a two-arm, parallel-group randomised controlled trial with a 1:1 allocation ratio. Following recruitment, participants will be randomly allocated to either the intervention or control group. The intervention group will receive personalised brain health coaching through the AI-coach for 12 months, while the control group will receive standard care, followed by access to the AI-coach for the subsequent 6 months. This delayed introduction of the AI-coach allows comparison of intervention effectiveness at 6 months as the primary outcome timepoint, with secondary assessments at 12 months to evaluate sustainability of change and within-group effects
Eligibility
Sex/Gender
ALL
Age
40 Years to No maximum
Healthy volunteers
Yes
Inclusion criteria
* Community-dwelling adults * Aged 40 years and above at time of recruitment * Able to use a smartphone and smartphone applications * Able to provide written informed consent
Exclusion criteria
* Been diagnosed wotj dementia, neurocognitive disorder or any major psychiatric illness * Have severe sensory or physical disability precluding usage of the application * Terminal illness or expected survival of less than 12 months at point of recruitment
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Brain Care Score | 12 months | To evaluate whether the AI-powered Brain Health Coach, BLAZE, improves BCS in community-dwelling adults versus usual care. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| NHG Healthy Eating Score | 12 months | A self-reported questionnaire assessing dietary habits aligned with brain health recommendations. Scores range from 0 to 100, with higher scores indicating healthier dietary behaviours. Improvement in score reflects adoption of healthier eating patterns across domains targeted by the intervention. |
| EuroQol-5 (EQ-5D) | 12 months | A standardised measure of health-related quality of life assessing five dimensions: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. Each dimension is rated on a scale of 1 (no problems) to 5 (extreme problems/unable to perform), generating a health utility index score where higher values indicate better health-related quality of life. Participants also rate their overall health on a Visual Analogue Scale (VAS) from 0 (worst imaginable health) to 100 (best imaginable health). |
| Lawton Instrumental Activities of Daily Living (IADL) | 12 months | A self-reported measure assessing the ability to perform complex daily tasks such as managing finances, using transportation, preparing meals, and using the telephone. Scores range from 0 to 8, with higher scores indicating greater functional independence. This measure monitors maintenance of functional independence in community-dwelling adults over the course of the study. |
| Katz Index of Independence in Activities of Daily Living (ADL) | 12 months | A measure assessing independence in six basic daily activities: bathing, dressing, toileting, transferring, continence, and feeding. Each activity is scored as independent (1) or dependent (0), yielding a total score of 0 to 6, where higher scores indicate greater independence. This measure monitors for any unexpected decline in basic functional independence during the study period. |
| UCLA Loneliness Scale (UCLA-3) | 12 months | A 3-item self-reported measure assessing subjective feelings of loneliness and social isolation. Each item is rated on a scale of 1 (hardly ever) to 3 (often), with total scores ranging from 3 to 9. Higher scores indicate greater loneliness. A reduction in score reflects improved social connectedness targeted through the intervention's social engagement recommendations. |
| Health Status (Self-Rated Health) | 12 months | A single-item measure asking participants to rate their overall health on a scale of 1 (excellent) to 5 (poor), or alternatively using a 5-point scale from poor to excellent depending on the version used. Higher scores on the favourable end of the scale indicate better perceived health status. This measure captures overall perceived wellbeing as a downstream indicator of sustained lifestyle behaviour change. |
| Lubben Social Network Scale (LSNS-6) | 12 months | A 6-item self-reported measure assessing the size and frequency of a participant's social network, including family and friend contacts. Scores range from 0 to 30, with higher scores indicating stronger social networks. A score below 12 is considered indicative of social isolation risk. Improvement in score reflects strengthened social connections targeted through the intervention's social health domain recommendations. |
| Chair Stand Test | 12 months | An objective performance-based measure of lower body strength and physical function. Participants are timed completing five sit-to-stand repetitions from a chair without using their arms. Faster completion times indicate better lower body strength and physical function. Improvement in performance reflects increased physical activity and strength targeted through the intervention's physical health recommendations. |
| Grip Strength | 12 months | An objective measure of overall muscular strength assessed using a handheld dynamometer. Results are recorded in kilograms, with higher values indicating greater muscular strength. Improvement in grip strength reflects sustained engagement in physical activity behaviours targeted through the intervention. |
| Adapted Dementia Knowledge and Attitudes (14-item) | 12 months | A 14-item adapted questionnaire assessing knowledge of dementia risk factors, symptoms, and prevention strategies, as well as attitudes towards dementia. Six items were adapted from the validated Dementia Knowledge and Attitudes Scale (DKAS) and ten items from a comparative analysis of dementia health literacy in Singapore. Higher scores indicate greater dementia knowledge and more positive attitudes towards dementia. |
| Partners in Health Scale (PIH) | 12 months | A 12-item self-reported measure assessing self-management capacity across four domains: knowledge of condition, adherence to treatment, recognition and management of symptoms, and coping. Each item is rated on a scale of 0 to 8, with total scores ranging from 0 to 96. Higher scores indicate greater self-management capacity. Improvement in score reflects the intervention's coaching module supporting independent brain health management. |
| Rapid Assessment of Physical Activity (RAPA) | 12 months | A self-reported questionnaire assessing aerobic physical activity levels and strength and flexibility behaviours. The aerobic component is scored from 1 to 7, where scores of 1-4 indicate sub-optimal activity levels and scores of 5-7 indicate recommended levels of moderate to vigorous physical activity. The strength and flexibility component is scored from 0 to 2. Higher scores across both components indicate more active lifestyle behaviours. Improvement reflects the intervention's physical activity nudges and goal-setting support. |
| Montreal Cognitive Assessment (MoCA) | 12 months | A brief 30-item cognitive screening tool assessing multiple cognitive domains including memory, attention, language, visuospatial ability, and orientation. Scores range from 0 to 30, with higher scores indicating better cognitive function. A score of 26 or above is considered within the normal range. This measure provides a baseline cognitive profile and enables monitoring of any meaningful change in global cognitive function during the study period. |
Countries
Singapore
Contacts
CONTACTDr Li Feng Tan
Outcome results
None listed