Data from "LIFETRACE: A Longitudinal Multimodal Dataset on Daily Physical Activity, Well-Being, and Habits"

Multimodal longitudinal dataset capturing anthropometrics, blood biomarkers, self-reported ecological momentary assessments, questionnaires, and wearable-derived activity traces from adult participants enrolled in a 14-week walking intervention study. Collection period: January–April 2022 (14 weeks). Participants: 88 adults (59 in daily walking protocol, 29 controls). Study sites: Università della Svizzera italiana (Switzerland) and Alexandru Ioan Cuza University (Romania). Access status: Restricted. Data can be shared upon reasonable request for academic, non-commercial research. Available Files (upon request approval) The following files are included in a .zip archive, which will be provided once your access request has been approved. CODEBOOK.md: Data dictionary with variable descriptions and coding schemes. anthropometric.csv: Anthropometric and body composition measures at M1, M2, M3 — includes BMI, body fat, muscle mass, metabolic age. blood.csv: Blood biomarkers and waist circumference at M1, M2, M3 — includes cholesterol, triglycerides, fasting context. emas.csv: Ecological Momentary Assessment responses (daily) — walking adherence, wellbeing, sleep, emotions. extra_variables.csv: Baseline demographics, health history, environment — lifestyle, comorbidities, ALPHA, IFIS scores. questionnaires.csv: Standardised psychological questionnaire scores — mental wellbeing, personality, LOT-R, PWB. wearables/activity.csv: Daily wearable activity summary — steps, distance, run distance, calories. wearables/sport.csv: Wearable sport-session metrics — sport duration, pace, distance, calories. Code Availability The code used for data preprocessing, analysis, machine learning experiments, and figure generation is available in the GitHub repository connected to this dataset: https://github.com/USI-PCC/LIFETRACE. Intended Uses Investigate determinants of long-term physical activity habit formation. Explore relationships between environmental context, physiological markers, and wellbeing. Benchmark machine learning models for predicting wellbeing metrics from multimodal signals. Data Collection Protocol Summary Participants in the intervention groups received daily prompts to engage in walking sessions and complete EMA surveys covering affect, context, and sleep. Wearable devices recorded activity and sport-specific metrics. Anthropometric and blood measurements were acquired during three laboratory visits (baseline, mid-study, post-study). Baseline questionnaires captured socio-demographic, psychological, and environmental context variables. Access Instructions To obtain access to the dataset files, please email both of the following addresses with a short description of your research project: Francesco Bombassei De Bona — [e-mail] Georgiana Juravle — [e-mail] Access will be granted for academic, non-commercial purposes and may require signing a data use agreement to ensure participant privacy. Citation When citing the dataset, please use both the following references: Francesco Bombassei De Bona, Ioana Andreea Câmpanu, Marc Langheinrich, Martin Gjoreski, and Georgiana Juravle. 2025. Data from "LIFETRACE: A Longitudinal Multimodal Dataset on Daily Physical Activity, Well-Being, and Habits" [Data set]. Zenodo. https://doi.org/10.5281/zenodo.17249917 Francesco Bombassei De Bona, Ioana Andreea Câmpanu, Marc Langheinrich, Martin Gjoreski, and Georgiana Juravle. 2025. LIFETRACE: A Longitudinal Multimodal Dataset on Daily Physical Activity, Well-Being, and Habits. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol. 9, 4, Article 161 (December 2025), 33 pages. https://doi.org/10.1145/3770676