Personalized Digital Training for COGnitive FITness in Mild Cognitive Impairment

The COG-FIT study is a pragmatic, single-blind, randomized controlled trial designed to assess the effectiveness of RICORDO, a CE-marked digital cognitive rehabilitation intervention, in improving patient activation and autonomy in individuals with Mild Cognitive Impairment (MCI) and Subjective Cognitive Decline (SCD). The study is sponsored by the University of Modena and Reggio Emilia and funded under the Italian National Recovery and Resilience Plan (PNC0000007 - FIT4MEDROB). It is a post-marketing clinical investigation of a certified medical device software, conducted in accordance with EU Regulation 2017/745. Participants (n=100) will be recruited from two outpatient centres in Modena and Milan, Italy. Eligible individuals will be randomly assigned (1:1) to either: RICORDO group: 15 home-based sessions (3/week, 45 minutes each) using a personalized digital training program delivered via tablet, guided by adaptive algorithms based on the patient's MoCA profile, baseline clinical information and performance over time. Active control group (S.A.M.B.A.): 15 sessions with video and printed materials on lifestyle domains (Social interaction, Nutrition, Physical activity, Psychological well-being, Environment), matched for duration and frequency. Primary and secondary outcome measures will be collected at baseline (T0) and after 5 weeks (T1) by blinded assessors. The primary endpoint is the change in Patient Activation Measure (PAM). Secondary endpoints include cognitive function (MoCA, CRIq), functional status (WHODAS 2.0), quality of life (EQ-5D-5L), and neuropsychiatric symptoms (STAI-Y, BDI-II, MBI-C). Usability (SUS) and technology acceptance (TAM) will be assessed in the RICORDO group at T1. Participants' adherence will be tracked digitally (RICORDO) or via logbooks (S.A.M.B.A.), with weekly phone support. Data will be pseudonymized and entered into REDCap servers hosted in Italy, which are managed in accordance with GDPR and national privacy laws. The study is powered to detect a mean PAM difference of 8.8 points (SD = 12) based on previous trials. With α = 0.05, power = 90%, and a 20% dropout rate, the total sample size is set at 100. Statistical analyses will be conducted on an intention-to-treat basis, with imputation of missing data and sensitivity checks. No invasive procedures or pharmacological interventions are planned. All participants will provide informed consent. Risk is minimal due to the non-invasive nature of the intervention, and adverse events will be closely monitored. The total duration of the study is 12 months, with 9 months of recruitment and a 5-week intervention per participant. Results will be disseminated through open-access publications, public events, and the project's website.