PhysAgeNet Deliverable D2.4 Demonstrator (prototype) of the repository including real data from technology-assisted PA interventions

CA20104 – Network on evidence-based physical activity in old age (PhysAgeNet) Deliverable D2.4 Demonstrator (prototype) of the repository including real data from technology-assisted PA interventions Contributors Working Group 2 Carl-Philipp Jansen (WG2 co-leader), Tiia Kekäläinen (WG2 leader), Salvatore Tedesco (WG2 member), Orgesa Qipo (WG2 member), Ivan Bautmans (WG2 member), Erja Portegijs (WG2 member, former WG2 leader) Table of Contents 1. Introduction ................................................................................................................................................................................. 3 2. T6: Create a structure for variables composing the open repository ................................................... 4 2.1. Objective ................................................................................................................................................................................. 4 2.2. Approach ........................................................................................................................................................................... 4 3. T7: Find suitable platform from existing open repositories ........................................................................ 9 3.1. Objective ................................................................................................................................................................................. 9 3.2. Evaluation Criteria ...................................................................................................................................................... 9 3.3. Justification for Selection ...................................................................................................................................... 9 4. T8: Prepare actual data delivery with small projects of technology-assisted physical activity interventions in older people .......................................................................................................................................... 11 4.1. Objective ........................................................................................................................................................................... 11 4.2. Implementation ........................................................................................................................................................... 11 5. T9: Data-delivery in open repository completed ................................................................................................ 11 6. References ............................................................................................................................................................................ 12 1. INTRODUCTION The provision of data for scientific research is gaining importance – but also in complexity. In the field of health research in particular, effective data sharing is essential to promote collaboration, increase efficiency, and drive innovation. This deliverable focuses on the practical implementation of a repository designed to support the structured and legally compliant provision of research data. The goal is to create a foundation that enables: • stronger cross-project collaboration, • more comprehensive analysis of large datasets, and • improved patient outcomes through data-driven innovation. This deliverable reports on the following tasks that are listed under D2.4: T6: Create a structure for variables composing the open repository; T7: Find suitable platform from existing open repositories; T8: Prepare actual data delivery with small projects of technology-assisted physical activity interventions in older people; T9: Data-delivery in open repository completed. 2. T6: CREATE A STRUCTURE FOR VARIABLES COMPOSING THE OPEN REPOSITORY 2.1. Objective The goal of this task is to define a minimal dataset structure for aggregated data reporting in studies on technology-assisted physical activity (PA) interventions in older adults. 2.2. Approach Review data extraction sheets from the systematic reviews listed. These reviews include diverse endpoints like neuromuscular biomarkers, physical performance, inflammatory markers, and behavioural factors. From this, define a standardised variable list across studies. This structure is based on a synthesis of the parameters typically reported in the systematic reviews referenced, aligned with FAIR data principles (Findable, Accessible, Interoperable, Reusable), and tailored for future meta-analyses and evidence synthesis. 1. Study-Level Metadata Variable Name Type Description Study_ID String Unique identifier (e.g. FirstAuthor_Year) Year_of_Publication Integer Actual year of publication DOI String Digital Object Identifier of the publication Title String Full study title Authors String/List Names of first and last author (or full list if needed) Country Categorical Country where the study was conducted Study_Design Categorical RCT / Cohort / Pilot / Feasibility study / Case series Recruitment_Source Categorical Community-dwelling / Assisted living / Clinical Ethical_Approval Binary Was ethical approval reported? Yes/No Conflict_of_Interest Binary Was a COI declaration provided? Adverse_events Binary Were AEs reported? Yes/No 2. Participant Characteristics Variable Name Type Description Sample_Size_Total Integer Total number of participants Mean_Age Integer Mean age ± SD of participants Female_Percent Integer Percent of sample that is female Clinical_sample Binary Clinical sample: yes or no Clinical_diagnosis String Primary diagnosis relevant for mobility, e.g. Parkinson’s, MS, Frailty, … Inclusion_Criteria String Textual description or coded field (e.g., age ≥ 65, MMSE > 24) Exclusion_Criteria String Common exclusion: cognitive impairment, pacemaker, severe visual impairment 3. Intervention Details Variable Name Type Description Exercise_Type Categorical e.g., running, walking, balance, resistance training, power training, multimodal, multicomponent,… Intervention_Type String e.g., Exergame, Wearable PA tracker, Robotics, e-Coaching, Mobile App, Online Training,… Tech_Device String Device used for intervention delivery, e.g., Tablet, Smartphone, PC, Gaming console Tech_Spec String Brand/model of the device if applicable (e.g., Fitbit Inspire, Nintendo Wii, Windows PC, …) Intervt_Duration_weeks Integer Total length of the intervention in weeks Sessions_Per_Week Integer Frequency of the sessions Session_Length_Minutes Integer Duration of a single session Session_Intensity Integer Intensity of exercise(s) Supervision_Type Categorical Fully supervised / Partially supervised / Remoteonly / Not supervised Control_Group Categorical Type of control group: Active controls / Inactive controls / Waiting controls / No controls 4. Outcome Variables – Aggregated Data Variable Units Notes Primary_outcome String Name of primary outcome variable Prim_outc_bl Integer Baseline value: Median and IQR or Mean ± SD Prim_outc_post Integer Post value: Median and IQR or Mean ± SD Further (secondary) outcome variables from then on can be grouped thematically based on the type of intervention and review focus. 4a. Physical Performance Metrics The following are the most often used measures of intrinsic capacity. These are not mandatory, but are suggested to be used for compatibility across studies. Other measures can be added, of course. Variable Units Notes Gait_Speed_Pre Integer Mean ± SD [m/s] baseline value Gait_Speed_Post Integer Mean ± SD [m/s] post value SPPB_Score_Pre Integer Short Physical Performance Battery baseline measurement value [Score] SPPB_Score_Post Integer Short Physical Performance Battery post measurement value [Score] Handgrip_Strength Integer Dynamometry baseline [kg] baseline value TUG_Time_Pre Integer Timed Up & Go Test baseline value [sec] TUG_Time_Post Integer Timed Up & Go Test post value [sec] … continued … use important physical performance outcomes as necessary 4b. Biomarkers Variable Type Notes Biomarker_use Binary Was there a measurement of biomarkers? Yes/no Biomarker_1 String Name of biomarker 1 Biomarker_1_Pre Integer Baseline value [unit] of biomarker 1 Biomarker_1_Post Integer Post value [unit] of biomarker 1 Biomarker_2 String Name of biomarker 1 … continued … … 4c. Quality of life / Motivation / Adherence Inclusion of measures of adherence is highly recommended. If measures of adherence, motivation, and quality of life are available, these can be reported under this category. Several measures of adherence can be reported (e.g., attendance, completion, …); the list would have to be adapted accordingly Variable Type Notes Adherence_Rate_Technol Integer Percentage of sessions completed using the technology Adherence_Rate_Exercise Integer Percentage of expected exercises completed (if possibly differing from Technology use) Dropout_Rate_Percent Integer % of participants lost to follow-up Motivation_Scale_Pre Integer e.g., Intrinsic Motivation Inventory QoL_Score_Post Integer SF-36 / WHOQOL-BREF, etc. 5. Quality/Metadata Flags (Optional) Report if available. Variable Type Description Risk_Of_Bias_Score Integer/Categorical e.g., ROB2 tool GRADE_Certainty_Level Categorical High / Moderate / Low Funding_Source String Public / Private / Mixed Conflict_of_Interest_YN Binary Yes / No AEs_Number Integer Number of adverse events 3. T7: FIND SUITABLE PLATFORM FROM EXISTING OPEN REPOSITORIES 3.1. Objective The objective of this task is to identify a suitable platform from the landscape of existing open data repositories that aligns with the principles of Open Science, supports long-term data preservation, and complies with European legal and ethical standards, notably the General Data Protection Regulation (GDPR). The selected platform should serve the needs of researchers within the COST Action and beyond, and enable compliant and sustainable sharing of research data outputs. 3.2. Evaluation Criteria In line with the requirements of COST Actions and broader EU funding schemes (e.g. Horizon Europe), the following criteria were defined for evaluating suitable repositories: • GDPR compliance and data governance • Support for FAIR data principles (Findable, Accessible, Interoperable, Reusable) • Open access and inclusivity • Cost structure and sustainability • Technical infrastructure and usability • Support for citation, versioning, and persistent identifiers (DOIs) • Compatibility with EU funders’ mandates and Open Science policies Selected Repository: Zenodo Following a comparative review of multiple repository platforms (including Figshare, Dryad, Mendeley Data, and institutional options), Zenodo has been identified as a highly suitable platform for data sharing within the COST Action framework. 3.3. Justification for Selection 1. GDPR Compliance and Legal Reliability Zenodo is hosted by CERN and operated under the OpenAIRE initiative, which ensures full compliance with the EU General Data Protection Regulation (GDPR). This includes transparent data handling policies, anonymization capabilities, and robust user consent management, making it a secure choice for projects involving sensitive or personal data.