Introduction Cognitive frailty is common in community-dwelling older people and is an at-risk state for adverse health outcomes such as dementia, dependency, and mortality. Fortunately, cognitive frailty is reversible, with a higher probability of reversibility at earlier stages. Physical activity is known to play a significant role in reversing cognitive frailty; its effect is moderated by intensity and sustainability. However, physical inactivity is very common in older people and is one of the key phenotypical characteristics of cognitive frailty. Moderate to vigorous physical activity (MVPA) can reduce the risk of worsening cognitive frailty. Brisk walking is a simple form of exercise that can be practised by community-dwelling older people every day to boost their physical activity to or above a moderate intensity level. Conventional behavioural change interventions (CBCIs) have been shown to effectively engage sedentary older people in physical activity, but their effect size is small. The use of e-health methods that adopt existing and popular e-platforms (e.g., Samsung Health and WhatsApp) to promote specific behaviours (e.g., regular brisk walking) in specific groups (e.g., older people with cognitive frailty) is an innovative, practically feasible and theoretically sound method of increasing MVPA. However, the relative effectiveness of e-health interventions and CBCIs in vulnerable groups (i.e., older people with cognitive frailty) is unknown.

Objectives The objectives of this study are to compare the effectiveness of an e-health intervention and a conventional behavioural change intervention in older people with cognitive frailty in improving 1) moderate-to-vigorous physical activity, 2) reducing cognitive frailty, 3) improving cognitive function, 4) improving walking speed, 5) improving functional fitness, and 6) improving physical activity motivation

Methods A single-blinded, two-parallel-group, non-inferiority, randomised controlled trial will be conducted in a community setting. Subjects will be recruited from five elderly community centres in Hong Kong. The eligibility criteria will be as follows: (1) aged ≥ 60, (2) cognitively frail, (3) physically inactive and (4) possessing a smartphone. The participants in the intervention group will receive an e-health intervention. Those in the control group will receive a CBCI. Each intervention will last for 14 weeks. The outcomes will be MVPA min/week (primary), as measured by a wrist-worn ActiGraph; cognitive frailty, as measured by an ordinal scale; cognitive function, as measured by the Montreal Cognitive Assessment; and frailty, as measured by the Fried frailty phenotype (FFP). The outcomes will be assessed at T0 (baseline), T1 (immediately post-intervention) and T2 (6 months post-intervention). The investigators plan to recruit 192 subjects. Permuted block randomisation with randomly selected block sizes in a ratio of 1:1 will be used. Only the outcome assessors will be blinded. Four generalised estimating equations will be used to test the effects of the interventions on the four outcomes, which will be the dependent variables. The independent variables will be group, time and [group] × [time]. The level of significance will be set at 0.05.

Significance If the e-health intervention proves to be more effective and sustainable than the CBCI, There will be evidence suggesting that e-health interventions can replace CBCIs in promoting MVPA and treating cognitive frailty in older people in community settings. Further studies could then examine the potential role of e-health interventions in delaying the onset of dementia and dependency.

Inclusion Criteria:

• 1) aged ≥ 60 years, because the WHO defines older people in Asia as those aged 60 or older;
• 2) presence of cognitive frailty, defined as the co-existence of MCI and frailty at either the pre-frail or the frail level;
• 3) physical inactivity, defined as < 150 min of MVPA every week in the 4 weeks preceding the study, as confirmed by a Rapid Assessment of Physical Activity (RAPA) score of ≤ 4 (i.e., regularly underactive);
• 4) possession of a smartphone on which Samsung Health and WhatsApp can be installed for > 6 months; and
• 5) being a Cantonese speaker, defined as possessing the ability to communicate in Cantonese.

Exclusion Criteria:

• 1) impaired mobility caused by conditions that require medication (e.g., severe arthritis) and inability to walk briskly outdoors, as defined by a modified Functional Ambulatory Classification score of < 7);
• 2) depressive symptomatology, as defined by a Geriatric Depression Scale (GDS) score of ≥ 8;67
• 3) having probable dementia (i.e., a Montreal Cognitive Assessment [MoCA] score of < 20 or Clinical Dementia Rating [CDR] score of ≥ 1),
• 4) being physically unfit because of chronic illness, as defined by a Physical Activity Readiness Questionnaire (PAR-Q) score of ≥ 1, and not having been examined by a medical doctor and certified as physically fit to participate in brisk walking;
• 5) not enrolled in any other interventions to promote physical activity or training

Data will be provided upon other researchers' requests